{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T07:55:30Z","timestamp":1792050930182,"version":"4.4.0"},"posted":{"date-parts":[[2026]]},"group-title":"SSRN","reference-count":0,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"<jats:p>This study examines whether machine learning improves asset-class return forecasts, whether crosscountry information pooling further improves accuracy, and whether these gains also lead to better portfolio decisions. Using annual data for 16 countries covering equities, long-term bonds, bills, and housing from 1870 to 2020, we combine historical returns with macroeconomic and financial variables in a recursive forecasting framework. Pooled machine-learning forecasts significantly outperform both historical-mean forecasts and country-specific ML. These gains translate into higher Sharpe ratios for both international investors allocating across countries and domestic investors investing only within their home markets. Pooled ML also outperforms simple allocation rules such as equal weighting. The passive benchmark based on capital-weighted aggregate asset returns remains competitive in the baseline analysis, but pooled ML performs better when the universe is restricted to liquid asset classes or portfolio limits are relaxed. The results extend the machine-learning literature on long-run multi-asset allocation and highlight the value of pooling information across markets.<\/jats:p>","DOI":"10.2139\/ssrn.7610204","type":"posted-content","created":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T07:42:17Z","timestamp":1792050137000},"source":"Crossref","is-referenced-by-count":0,"title":["Multi-Asset Portfolio Selection with Machine Learning: Long-Run International Evidence"],"prefix":"10.2139","author":[{"given":"Ray","family":"Friedenberger","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-3183-7062","authenticated-orcid":true,"given":"Felix","family":"Reichenbach","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","container-title":[],"original-title":[],"deposited":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T07:42:17Z","timestamp":1792050137000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ssrn.com\/abstract=7610204"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":0,"URL":"https:\/\/doi.org\/10.2139\/ssrn.7610204","relation":{},"subject":[],"published":{"date-parts":[[2026]]},"subtype":"preprint"}}